HIV counseling and testing (HCT) is a highly cost-effective intervention for increasing serostatus awareness, a point of entry into HIV care and treatment, and an important means of primary and secondary HIV prevention. Public health officials have called for dramatic increases in HIV testing to achieve an HIV-free generation. However, testing rates are plateauing, and repeat testing rates among those with ongoing risk remain low. Novel approaches are needed to increase the uptake of HCT, especially among high-risk groups. Evaluations of HIV testing uptake in sub-Saharan Africa, including our own, have typically focused on the acceptability of specific venue-based testing options. These narrow assessments do not probe the potential diversity in testing preferences of target populations and cannot characterize testing options that will maximize uptake among heterogeneous groups. Our HIV Testing Preferences in Tanzania study (Multi-PI Ostermann, Thielman, 5R21MH096631) demonstrated that the Discrete Choice Experiment (DCE), a form of stated preference survey research, is a robust tool for identifying which characteristics of HIV testing options are most preferred by different populations and which tradeoffs individuals make in evaluating testing options. Building on the R21, a decade of productive HCT research in this region, and strong relationships with policy makers and implementers, we propose to test the hypothesis that offering HIV testing options that reflect the specific preferences of high-risk populations will significantly increase rates of testing among these groups. In Aim 1 we will characterize the HIV testing preferences of two high risk populations. DCEs with samples of 200 female barworkers and 200 male mountain porters, two high-risk groups in Moshi, Tanzania, will be used to identify the relative importance of diverse HIV testing characteristics. In Aim 2, we will compare existing testing options to the preferences of each high-risk group and identify feasible, more preferred testing alternatives, taking into consideration that one-size-fits-all approaches cannot meet the heterogeneous preferences of these populations. In Aim 3 we will conduct a pragmatic randomized controlled trial to evaluate the effect of a preference-based HIV counseling and testing (PB-HCT) intervention on testing uptake. Representative samples of 600 female barworkers and 600 male mountain porters will be randomized into two groups. Group A participants will be offered the set of predicted more- preferred PB-HCT options identified in Aim 2 and a default alternative; Group B participants will be offered predicted less-preferred alternatives and a default alternative. Uptake of testing and the cost effectiveness of a preference-based HCT intervention compared to a default testing offer will be assessed after 6 and 12 months. The proposed study will identify the effect of an HIV-testing intervention designed using a novel, preference- based strategy. If successful, this work will demonstrate the utility of DCEs as a tool to replace the costly practice of iteratively implementing narrowly focused interventions with a structured approach for developing interventions matched to the specific preferences of intended target populations.